基于凸优化技术的改进型卡尔曼滤波算法

Improved Kalman Filter Based on Convex Optimization Technique

为了能够在高斯噪声和稀疏噪声混合情况下对目标进行准确跟踪,提出基于凸优化的改进型卡尔曼目标跟踪算法。改进后的方法以传统卡尔曼滤波方法为基础,结合凸优化技术,从最大后验估计理论和贝叶斯理论的角度构建目标跟踪的优化问题,将噪声统计特性作为先验约束引入优化过程中,实现在高斯噪声和稀疏噪声混合情况下对目标的准确跟踪。仿真实验结果证明该方法的可行性和有效性。

In order to track objects under Gauss noise and sparse noise conditions with high accuracy, an improved Kalman object tracking method based on convex optimization is proposed. By convex optimization technique, the improved method makes use of statistics feature of various kinds of noise to achieve robustness against Gauss noise and sparse noise in the perspective of maximum posterior estimation theory and Bayesian theory. The experiment results show the feasibility and effectiveness of the proposed method.